The effects of a competing message on the speech comfortable loudness level (CLL) was compared for two instructional sets in 2 conditions of quiet and in the presence of a four-talker babble competing message. Kopra and Blosser's (24) instructions were selected to sample the middle of the CLL range, whereas Lucker et al.'s (27) instructions sampled the upper limit of the CLL range. Two groups of 12 normal-hearing subjects were tested using a method of adjustment. The results indicated no significant differences between the initial and final CLL's in quiet. However, data for each instructional set showed that the CLL's obtained with Lucker et al.'s phraseology (97.8 dB SPL) was significantly higher than with Kopra and Blosser's instructions (75.2 dB SPL). The four-talker babble served to raise the CLL, although less effect was observed on the upper limit of the CLL range than for the middle portion of this range. The clinical implications of these findings are discussed.